THE BENEFITS OF REVERSE ENGINEERING FOR ENSURING PIPELINE INTÉGRITY
Author: Jérôme-Alexandre Lavoie, Product Manager
ABSTRACT
Today, now more than ever before, mounting public concern over pipeline safety has
incited many companies to turn to emerging methodologies and technologies to better
monitor and assess pipeline degradation. The ultimate goal is to predict the evolution of
pipeline degradation so as to deploy preventative measures in a timely fashion—before
interventions become too costly and the safety of citizens is compromised.
Reverse engineering is a new tool to help pipeline integrity assessment of complex or
combined damages. This paper will describe the various roles reverse engineering plays
as well as the different reverse-engineering approaches that can be used for pipeline
assessment. This paper will also cover the most suitable 3D technologies—and their
distinct advantages—for ensuring optimal assessments.
www.creaform3d.com Page 2 of 12
INTRODUCTION
Reverse engineering is the process that identifies an object, a device, or a system’s
technological properties by performing a comprehensive analysis of its structure,
functions and operations. In mechanical engineering, this process aims to create a
virtual 3D model from an existing physical object in order to duplicate or enhance it.
They are many reasons to use reverse engineering of physical objects. For example, the
reverse engineering process will be used if the original design is not supported by
sufficient or adequate documentation or if the original CAD model is not appropriate to
support modifications and/or standard production methods. In some cases, the original
manufacturer no longer exists or manufactures a product—yet some requirements
remain for that product. Reverse engineering would then be the ideal solution..
In order to create a 3D model of the object, the object must be measured using 3D
scanning technologies (CMM, laser scanners, structured light digitizers, etc.). Once the
scanning is completed, it is usually possible to rebuild the 3D model using 3D CAD,
CAM, CAE or other software.
HANDYSCAN 3D BY CREAFORM
HandySCAN 3DTM is a handheld scanner manufactured
by Creaform. It is a very portable and versatile self-
positioning 3D scanner. It uses Creaform’s positioning
targets to reference its position from the object to scan.
This device’s main characteristic is its portability; it
refers directly to the object with targets. It can easily be
carried to the object instead of bringing the object to the
scanner.
Once the object and the scanner’s position have been
located with targets, the surface acquisition is completed
via the camera. The camera detects laser lines that
cross each other and are projected onto the surface. As
the surface is swept over by the laser, data is recorded
based on the triangulated position. The output file format
is a STL file.
STL (stereo-lithography) is a file format native to the stereo-lithography CAD software,
widely used in 3D printing and CAD industries. This file format is supported by many
other software packages. STL files only contain the surface’s geometry of a three-
dimensional object without any color, texture or other common CAD model attributes.
They also contain a raw unstructured triangulated surface of the unit normal and vertices
(ordered by the right-hand rule) using a three-dimensional Cartesian coordinate system.
www.creaform3d.com Page 3 of 12
Using STL files for reverse engineering has several advantages. They can:
Recognize features intuitively
Distinguish front and back faces
Evaluate curvature
Measure volume
Accurately measure alignment between scan data
Be used for visualization and documentation
REVERSE ENGINEERING APPLIED TO PIPELINE INTEGRITY
These days, companies are aware of public concern about pipeline safety and are
turning to new technologies to increase public confidence towards pipeline installations.
A constant monitoring of the installation is performed with tools capable of detecting
anomalies such as loss of wall thickness and mechanical damage and other.
Pipeline integrity assessment is a process that falls under the responsibility of Quality
Control (QC) and metrology. As QC is a common process in part manufacturing, pipeline
assessment is key in the pipeline integrity management program of asset owner.
Recent technological breakthroughs from Creaform allow easy access to important
information, which can help pipeline operators to increase the safety of their networks.
PipecheckTM software, Creaform’s proprietary solution for pipeline integrity assessment,
is the only software solution on the market to combine the power of reverse engineering
and QC methodology for pipeline integrity assessment.
There are several definitions of quality depending on the application. Manufacturing-
based definitions associate quality with the conformance of a product to its
specifications. QC) is a process that aims to review the quality of all factors involved in
the production process and part life cycle.
Metrology is the science of measurement. In metrology, precision refers to the
dispersion of measurements. The measurement error (the mean) can be close to zero
even if the system is very not precise (it nevertheless must have a good trueness). In
other words, the less the measurement data is scattered, the more the equipment is
precise. A formal definition of precision is: closeness of agreement between indications
of measured quantity values obtained by replicate measurements on the same or similar
objects and under specified conditions.
The word trueness gives information on the difference between the mean of
measurements and the real dimension, regardless of dispersion. In other words, the
www.creaform3d.com Page 4 of 12
more the mean of measurements is close to the nominal value, the more the equipment
has good trueness. A formal definition of trueness is: closeness of agreement between
the average of an infinite number of replicate measured quantity values and reference
quantity value.
Accuracy is the conformity between scan data and reality. To evaluate the accuracy of
a measuring device, such as a laser scanner, the data acquired with the device should
be compared to the data acquired with a more accurate measurement tool (e.g.: a
coordinate measuring machine (CMM)). In addition, the measured item must be
normalized. A more formal definition of measurement accuracy is: the closeness of
agreement between a measured quantity value and a true quantity value.
Figure 1: Representation of accuracy, trueness and precision.
Uncertainty is described as the doubt about the validity of a measurement result. A
measurement never gives the true value; however, it is the best estimation of it. This
way, a measurement is only complete if it is accompanied by a statement of the
associated uncertainty. The following figure shows the main sources of uncertainty.
www.creaform3d.com Page 5 of 12
Figure 2: Representation of main sources of uncertainty.
Equipment performance is not the only thing to consider for highest accuracy possible.
Due to several potential sources of error, uncertainty is difficult to determine. It is a
decisive factor when it is required to make an informed inspection decision.
Applying reverse engineering concept to perform pipeline integrity assessment is an
interesting idea when performed properly. There are several existing reverse
engineering methods available, but not all of them are suitable for pipeline integrity
assessment.
Parametric Reverse Engineering
This method for reverse engineering is also known as “design intent.” The main goal of
using such techniques is to understand how the object was conceived. Some software
programs are dedicated to this application. They include actual conception tools with
typical geometrical shapes. Basically, the workflow consists of analyzing the several
parts composing the object and creating similar features, as one would do in CAD
software (extrusion, revolution, geometries, etc.) to match its global shape.
The physical object or a scanned 3D model is made up of approximate shapes.
Therefore, an object conceived in this reverse engineering process is never the exact
copy of the reference. As it is composed of perfect conception features, it is always
slightly different. The reference and the reverse engineering model need to be compared
to each other in order to evaluate their relative deviations. If the deviations are too high
and the tolerance is not respected, modifications may have to be applied to solve the
problem.
www.creaform3d.com Page 6 of 12
Freeform Reverse Engineering
With this method, the idea is not to understand the conception of an object, but only to
create a surface with the exact same shape. Using such techniques, it is possible to
efficiently extract accurate freeform surfaces from the shape of 3D scanned objects.
Freeform surfaces range anywhere from the shape of a rock formation to the body of a
car. In these cases, the STL model is used as a reference to build NURBS based on the
surface.
For this application, the scan surface needs to be as perfect as possible. Therefore, we
can distinguish two main steps for freeform reverse engineering: mesh
preparation/optimization and the actual surface reconstruction. These two stages are
strongly linked; a thorough preparation will ease the surface construction.
The mesh optimization sometimes involves multiple steps, including: alignments,
manifold, hole filling, assembly combination (merge operations), decimation, etc. All of
these concepts are highly proscribed when performing quality control since they can
modify the surface condition. The NURBS (Non-uniform rational B-spline) model to build
will therefore be performed on the direct output of the acquisition software.
Creaform’s Innovation
With this method, the idea is not to understand the conception of an object, but only to
create a surface with the exact same shape. Using such techniques, it is possible to
efficiently extract accurate freeform surfaces from the shape of 3D scanned objects.
Freeform surfaces range anywhere from the shape of a rock formation to the body of a
car. In these cases, the STL model is used as a reference to build NURBS based on the
surface.
Creaform’s breakthrough comes from the development of a hybrid technique between
the two approaches described above. When applied to pipeline integrity assessment,
especially for material loss located at the bottom of mechanical damage defects,
Creaform’s approach enables the extraction and extension of dent curvatures over
corroded areas. This approach requires the understanding of the dent shape prior to any
material loss—without surface optimization operations. Using the HandySCAN 700,
powered by the TRUaccuracy™ technology, and powerful surface fitting methods,
Pipecheck software offers a simple workflow to accurately measure corrosion depths at
the bottom of a dent.
The dent shape prior to corrosion is extracted using the surface of the dent. This surface
is user-defined: Pipecheck uses the dent surface area and removes what the user
manually defined as the corroded surface. These steps are performed manually to allow
the user total control over the exact area that is considered corroded.
www.creaform3d.com Page 7 of 12
Figure 3: Overall dent surface and corroded region representation
Creaform’s research and development team have considered many different methods
for the extrapolation of the material that falls under the corrosion damage. Considering
the pros and cons for each was an important part of the final decision to go for NURBS.
Some extrapolation techniques gave surfaces that were too rigid. For instance, looking
for the weighted mean of the edge of the corrosion and using that as the central point of
the extrapolation gave a surface that was too flat with no assurance of continuity along
the edges.
www.creaform3d.com Page 8 of 12
Figure 4: Material extension using a weighted mean
A second approach was to consider simple curve fitting using the main directions of the
pipe surface. The curves could be done in both directions independently and each would
have continuity along the edges. However, for this approach to reconcile the two
directions to form a single surface, interactions were needed between the curves; using
a simple mean of the two surfaces meant possibly losing some of the continuity in one
direction or another. It was better to consider parametric surfaces which would naturally
make the two directions interact.
Therefore, looking into parametric surfaces, two options seemed more interesting: Coon
patches and NURBS surfaces. Using a grid that follows the main directions of the pipe,
it’s possible to define parametric curves that follow the edges of the corrosion patches.
The curves can be used to define four independent surfaces. The Coon patch is the
interpolation of all four curves with the necessary continuity in all directions.
www.creaform3d.com Page 9 of 12
Figure 5: Parametric curves resulting in a Coon patch
The main drawback of using Coon patches is that this technique focuses on the
boundary information and the first derivatives of these boundaries. Therefore, the overall
dent surface information is ignored, which can lead to inaccurate reconstruction. NURBS
surface techniques using the dent surface surrounding the corrosion damage ensure
continuity throughout the entire interpolated surface. This technique brought the best
results during the validation technique.
Figure 6: NURBS surface reconstructed from the overall dent area without information from the corroded area.
www.creaform3d.com Page 10 of 12
CASE STUDY: FILL HOLES VS NURBS SURFACES
This section presents the comparison results between the two methods previously
mentioned: Coon patches (usually associated with hole filling methods) and NURBS
surfaces. The idea was to study both methods’ reaction and thus determining their
accuracy when reconstructing a surface. When applying these approaches to NDT,
specifically to corrosion analysis inside a dent, we focussed on how accurately the
deepest corrosion point can be extracted. Using four different samples, two specific
analyses were processed: one comparing the deepest corrosion points found and one
determining how human error influences each method’s accuracy. Prior to these
analyses, a data set was created following these steps:
a. Different corrosion patterns were numerically created on a real dent surface
scanned by a HandySCAN 3D.
Table 1: Sample data with corrosion
b. Nominal deepest corrosion points were found in each sample by creating a
deviation analysis from the original surface without corrosion. For Sample 3, a
deepest point for each of the six corrosion pits was found.
Table 2: Nominal deepest corrosion points for each sample
Samples Sample 1 Sample 2 Sample 3 Sample 4
Deepest point -2.195 -2.316 -4.088 -1.668 -1.406 -1.62 -0.972 -1.415 -3.465
c. Corrosion was selected and virtually deleted, thereby leaving a hole to be filled
by each method.
Table 3: Sample with corrosion virtually deleted
www.creaform3d.com Page 11 of 12
Analysis 1 – Deviations of deepest corrosion points for each method
Steps:
a. Dent surfaces were recreated from the samples with holes (corrosion deleted)
using each method. For Method 1 (Fill holes/Coon technique), three major
different software programs were used.
b. Deepest corrosion points were measured by creating a deviation analysis
between samples with corrosion and surfaces created by each method.
c. Deviations were computed based on the nominal deepest points that were found
earlier.
When comparing deviations, the NURBS Patches approach was better 6 times out of 8
on normal corrosion patterns. When less accurate, the NURBS Patches method slightly
overestimated the corrosion depths, giving a deeper measurement and thus creating a
more conservative reading. This behavior is better for NDT applications; when it is
always advisable to underestimate the pipeline’s remaining strength, rather than
overestimate it. NURBS Surfaces stability was also better: its mean deviation being
almost twice more accurate than the other approach. For abnormal corrosion shapes,
where the corroded area is very large in proportion to the dent surface, such as Sample
4, no method was accurate. It was impossible to determine the exact mechanical
damage shape prior to any corrosion.
NURBS Patches:
- More accurate 6 out of 8 times
- Average deviation is 0.019 mm compared to 0.031 mm
Table 4: Deviation (mm) of deepest corrosion points measurements for each method
Samples 1 2 3-1 3-2 3-3 3-4 3-5 3-6 3-7
Method 1 : Coon Patches
Software 1 0.004 0.072 -0.050 0.032 0.039 -0.012 0.032 -0.040 1.541
Software 2 -0.037 0.036 -0.018 0.056 0.058 0.027 0.022 -0.034 0.965
Software 3 0.000 0.002 0.014 0.039 0.036 0.031 -0.008 -0.041 2.011
Mean 0.014 0.037 0.027 0.042 0.044 0.023 0.021 0.038 1.506
Method 2 : NURBS Surfaces
Pipecheck -0.006 0.010 0.008 0.038 0.016 -0.003 -0.026 -0.045 2.367
Analysis 2 – Determining how human error influences each method In this second analysis, the hypothesis is that human error—and therefore corrosion
selection and deletion—will influence the accuracy of the methods. On the deepest
corrosion pit of Sample 3, three different selections from smaller to larger were deleted.
www.creaform3d.com Page 12 of 12
This replicates some inaccurate selections made by the user, were larger selection
passed the corrosion should lead to larger deviations.
Results indicate that both methods are affected by the selection, since the deviations get
larger on inaccurate Selection 2 & 3. The interesting fact is that the NURBS Surfaces
approach is much less affected by human error; its mean deviation on the three different
selections was more than twice better (0.035 mm VS 0.080 mm).
Table 5: Deviation (mm) of deepest corrosion points measurements for each selection and method
Selection 1 Selection 2 Selection 3
Method 1 : Coon Patches
0.027 0.139 -0.075
Method 2 : NURBS surfaces
0.008 0.069 0.028
CONCLUSION
As part of a comprehensive continuous improvement process, pipeline integrity
assessment has become mandatory for the industry. Reverse engineering can play an
important role in this process when done properly. Development and the appearance of
3D technologies on the market have definitely made available the use of this practice in
the pipeline industry. Although often seen as complicated, reverse engineering is now
greatly facilitated by the use of Creaform’s 3D accurate and reliable technologies and
software.